Method and apparatus for low signal to noise ratio instantaneous phase measurement

- Aerometrics, Inc.

The system and method of the present invention generates high resolution phase measurements without the high processing and memory overhead requirements found in prior art circuits and methods. Each signal sample (of N samples) is divided further into J segments each segment having K samples. The frequency is computed with respect to one J segment and is used in the phase measurement computations performed for the remaining segments. The phase measurements performed with respect to each J segment are then averaged to compute a high resolution phase measurement for the corresponding N segment.

Skip to:  ·  Claims  ·  References Cited  · Patent History  ·  Patent History

Claims

1. A method for measuring phase for N samples of a signal comprising the steps of:

determining a number of K samples within an N sample;
identifying a number of J segments within an N sample, each J segment composed of K samples;
computing a frequency measurement of the signal for one of the J segments;
computing a phase measurement of each J segment using the frequency computed for one of the J segments;
combining the phase measurement of each J segment to generate a combined phase measurement indicative of the phase measurement of the N sample of the signal.

2. The method as set forth in claim 1, wherein the step of determining the number of K samples comprises the step of determining K based upon the signal to noise ratio (SNR) of the signal.

3. The method as set forth in claim 1, wherein the step of determining the number of K samples comprises the step of determining K such that rms in phase measurement at the minimum expected signal to noise ratio (SNR) of the signal is less than.pi..

4. The method as set forth in claim 1, wherein the step of identifying the number of J segments comprises determining J according to the following equation: J=N/K.

5. The method as set forth in claim 1, wherein the step of computing the frequency comprises the steps of:

dividing the J segment of the signal into a plurality of sampled signals, each sampled signal sampled at a different sampling frequency:
performing a discrete Fourier transform (DFT) on each sampled signal; and
combining the results of the DFT sampled signals to determine the frequency.

6. The method as set forth in claim 1, wherein the step of computing the frequency comprises using zero crossing counting.

7. The method as set forth in claim 1, wherein the step of computing the frequency comprises using a fast Fourier transform (FFT).

8. The method as set forth in claim 1, wherein the step of computing a phase measurement for each J segment comprises determining the phase measurement according to the following:

If q.sub.i =1 then.phi..sub.i =.theta.
If q.sub.i =2 then.phi..sub.i =.pi.-.theta.
If q.sub.i =3 then.phi..sub.i =.pi.+.theta.
If q.sub.i =4 then.phi..sub.i =2.pi.-.theta.

10. A method for measuring phase for N samples of a signal comprising the steps of:

computing a frequency measurement of the signal for one of J segments within an N sample, each J segment composed of K samples;
computing a phase measurement of each J segment using the frequency computed for one of the J segments;
combining the phase measurement of each J segment to generate a combined phase measurement indicative of the phase measurement of the N sample of the signal.

11. The method as set forth in claim 10, wherein the number of K samples is determined based upon the signal to noise ratio (SNR) of the signal.

12. The method as set forth in claim 10, wherein the step of computing the frequency comprises the steps of:

dividing the J segment of the signal into a plurality of sampled signals, each sampled signal sampled at a different sampling frequency:
performing a discrete Fourier transform (DFT) on each sampled signal; and
combining the results of the DFT sampled signals to determine the frequency.

13. The method as set forth in claim 10, wherein the step of computing the frequency comprises using zero crossing counting.

14. The method as set forth in claim 10, wherein the step of computing the frequency comprises using a fast Fourier transform (FFT).

15. The method as set forth in claim 10, wherein the step of computing a phase measurement for each J segment comprises determining the phase measurement according to the following:

If q.sub.i =1 then.phi..sub.i =.theta.
If q.sub.i =2 then.phi..sub.i =.pi.-.theta.
If q.sub.i =3 then.phi..sub.i =.pi.+.theta.
If q.sub.i =4 then.phi..sub.i =2.pi.-.theta.

17. A circuit for measuring phase for N samples of a signal comprising:

at least one sampling circuit for sampling the signal;
a frequency measurement unit coupled to the at least one sampling circuit that computes a frequency measurement for one of J segments within an N sample, wherein each J segment is composed of K samples; and
a phase measurement unit coupled to the at least one sampling circuit and the frequency measurement unit to compute a phase measurement of each J segment using the frequency computed for one of the J segments and to combine the phase measurement of each J segment to generate a combined phase measurement indicative of the phase measurement of the N sample of the signal.

18. The circuit as set forth in claim 17, further comprising a burst detector coupled to the output of the at least one sampling circuit and to the frequency measurement unit and phase measurement unit, said burst detector detecting the presence of the signal and enables the operation of the frequency measurement unit and phase measurement unit.

19. The circuit as set forth in claim 17, wherein the frequency measurement unit measures the frequency by dividing the J segment of the signal into a plurality of sampled signals, each sampled signal sampled at a different sampling frequency, performing a discrete Fourier transform (DFT) on each sampled signal, and combining the results of the DFT sampled signals to determine the frequency.

20. The circuit as set forth in claim 17, wherein the frequency measurement unit comprises a zero crossing counting circuit.

21. The circuit as set forth in claim 17, wherein the frequency measurement unit measures the frequency by using a fast Fourier transform (FFT).

22. The circuit as set forth in claim 17, wherein the phase measurement unit comprises:

circuitry to compute a phase measurement for each J segment according to the following:
If q.sub.i =1 then.phi..sub.i =.theta.
If q.sub.i =2 then.phi..sub.i =.pi.-.theta.
If q.sub.i =3 then.phi..sub.i =.pi.+.theta.
If q.sub.i =4 then.phi..sub.i =2.pi.-.theta.

24. A circuit for measuring phase of N samples of a signal comprising the steps of:

at least one low resolution analog to digital converter (ADC) for sampling the signal;
a first logic computing a frequency measurement for one J segment of the sampled signal within an N sample, each J segment being composed of K samples;
a second logic for computing a phase measurement of each J segment using stored precomputed values; and
a third logic for combining the phase measurements of each J segment to generate the phase measurement for N samples.

25. The circuit as set forth in claim 24, wherein the signal is sampled using 1-bit ADC.

26. The circuit as set forth in claim 24, wherein the frequency is measured using the Discrete Fourier Transform (DFT).

27. The circuit as set forth in claim 24, further comprising memory and at least one latch, wherein the sampled signal is temporarily stored in memory and then latched by that at least one latch.

28. The circuit as set forth in claim 24, further comprising look-up tables (LUTs) used to store the precomputed values of several samples of the input signal with the correspondent values of the reference signal.

29. The circuit as set forth in claim 28, wherein the input to the LUTs are the sampled signal stored in the latches, the frequency and the phase of the reference signal.

30. The circuit as set forth in claim 28, wherein accumulators are used to sum the LUT's outputs to produce values of x and y where, ##EQU21## where r(i) and q(i) represent sampled data, k.sub.s represents an estimated signal frequency normalized to the sampling frequency, and K represents the number of samples within a segment.

31. The circuit as set forth in claim 30, wherein the second logic computes the phase.phi..sub.j of a segment j by computing tan.sup.-1 (x/y) and the quadrant of the phase measurement.

32. The circuit as set forth in claim 31, wherein the third logic computes the phase measurement of N samples by combining the phase measurements.phi..sub.j 's of valid segments.

33. A circuit for measuring phase over a large amplitude dynamic range comprising:

a plurality of sampling circuits for sampling the signal, each sampling circuit sampling different segments, each segment sampled at a sampling frequency fs;
a frequency measurement unit coupled to one sampling circuit of the plurality of sampling circuits;
a phase measurement unit to measure the phase of each segment and to combine the phase measurements to generate a phase measurement.

34. The circuit as set forth in claim 33, wherein at least one sampling circuit of the plurality of sampling circuits comprises a one bit analog to digital converter (ADC).

35. The circuit as set forth in claim 33, wherein each sampling circuit samples the signal for a segment using the sampling frequency of a previous segment delayed by.tau., where.tau. is a time delay.

36. The circuit as set forth in claim 33, wherein a pattern of the sampled data at fs/4, where s represents the sampling frequency, for each segment is different than that of other segments sampled.

37. The circuit as set forth in claim 33, wherein the sampling frequency for each segment is phase modulated.

38. The circuit as set forth in claim 33, wherein the frequency of one segment is computed using Fourier analysis.

39. The circuit as set forth in claim 33, wherein the phase of each segment is computed using Fourier analysis method.

40. The circuit as set forth in claim 33, wherein the phase measurement of N samples of the input signal are computed by combining the phase measurements of valid segments of N samples.

Referenced Cited
U.S. Patent Documents
4533873 August 6, 1985 Alcock
4700129 October 13, 1987 Yoshizawa et al.
4944036 July 24, 1990 Hyatt
5001724 March 19, 1991 Birgenbeier et al.
5019786 May 28, 1991 Fairley et al.
5179344 January 12, 1993 Najle et al.
5187719 February 16, 1993 Birgenheier et al.
5272447 December 21, 1993 Furuya et al.
5337014 August 9, 1994 Najle et al.
5359332 October 25, 1994 Allison et al.
5367538 November 22, 1994 LaRosa et al.
5432826 July 11, 1995 Rieder
5503159 April 2, 1996 Burton
Patent History
Patent number: 5808895
Type: Grant
Filed: May 31, 1996
Date of Patent: Sep 15, 1998
Assignee: Aerometrics, Inc. (Sunnyvale, CA)
Inventors: Khalid M. Ibrahim (San Jose, CA), William D. Bachalo (Los Altos Hills, CA)
Primary Examiner: James P. Trammell
Assistant Examiner: Cuong H. Nguyen
Law Firm: Blakely Sokoloff Taylor & Zafman, LLP
Application Number: 8/655,831
Classifications
Current U.S. Class: 364/484; 364/484; 364/485; 324/7647; 324/7677; 324/7682
International Classification: G01R 2500;